Quasi-model-independent search for new physics at large transverse momentum

Abstract

We apply a quasi-model-independent strategy ͑''SLEUTH''͒ to search for new high p T physics in Ϸ100 pb Ϫ1 of pp collisions at ͱsϭ1.8 TeV collected by the DØ experiment during 1992-1996 at the Fermilab Tevatron.Over 32 e X, Wϩjets-like, Zϩjets-like, and (l/␥)(l/␥)(l/␥)X exclusive final states are systematically analyzed for hints of physics beyond the standard model. Simultaneous sensitivity to a variety of models predicting new phenomena at the electroweak scale is demonstrated by testing the method on a particular signature in each set of final states. No evidence of new high p T physics is observed in the course of this search, and we find that 89% of an ensemble of hypothetical similar experimental runs would have produced a final state with a candidate signal more interesting than the most interesting observed in these data.

Anomaly detection with density estimation

“…• A series of signal model agnostic, but background model-dependent searches have been performed by D0 [10][11][12][13], H1 [14,15], CDF [16][17][18], CMS [19,20], and ATLAS [21][22][23]. All of these searches share essentially the same approach: they compared histograms of data to histograms of SM simulations and looked for discrepancies.…”

OPTIMASS: a package for the minimization of kinematic mass functions with constraints

“…Here M is the appropriate typical mass scale associated with the target mass function, its value should be chosen depending on the specific physics process at hand. In the current version of Optimass, the user is expected to provide the relevant fixed value of the scale 12 The specific method to set this quantity is where maxfcn denotes maximum number of function calls after which the Minuit minimization routine will be stopped even if it has not yet obtained satisfactory convergence to an acceptable minimum. The default value is maxfcn = 5000.…”

A model independent search for new physics in final states containing leptons at the D0 experiment

“…Once all of the event selection cuts and correction factors are determined in the MIS analysis packages, vista and sleuth consolidate this information and use it to quantify the agreement between the data and the standard model background. The sleuth algorithm was developed for DØ Run I [87,88,89] and later modified at the H1 experiment [90]. The vista algorithm was developed at CDF [91,28].…”